Dissection of the complex phenotype in cuticular mutants of Arabidopsis reveals a role of SERRATE as a mediator

Abstract

Mutations in LACERATA (LCR), FIDDLEHEAD (FDH), and BODYGUARD (BDG) cause a complex developmental syndrome that is consistent with an important role for these Arabidopsis genes in cuticle biogenesis. The genesis of their pleiotropic phenotypes is, however, poorly understood. We provide evidence that neither distorted depositions of cutin, nor deficiencies in the chemical composition of cuticular lipids, account for these features, instead suggesting that the mutants alleviate the functional disorder of the cuticle by reinforcing their defenses. To better understand how plants adapt to these mutations, we performed a genome-wide gene expression analysis. We found that apparent compensatory transcriptional responses in these mutants involve the induction of wax, cutin, cell wall, and defense genes. To gain greater insight into the mechanism by which cuticular mutations trigger this response in the plants, we performed an overlap meta-analysis, which is termed MASTA (MicroArray overlap Search Tool and Analysis), of differentially expressed genes. This suggested that different cell integrity pathways are recruited in cesA cellulose synthase and cuticular mutants. Using MASTA for an in silico suppressor/enhancer screen, we identified SERRATE (SE), which encodes a protein of RNA-processing multi-protein complexes, as a likely enhancer. In confirmation of this notion, the se lcr and se bdg double mutants eradicate severe leaf deformations as well as the organ fusions that are typical of lcr and bdg and other cuticular mutants. Also, lcr does not confer resistance to Botrytis cinerea in a se mutant background. We propose that there is a role for SERRATE-mediated RNA signaling in the cuticle integrity pathway.

Figure 1. Functional implications and consequences of mutations in LCR, BDG, and FDH.

(A) Three mutants, lcr, bdg and fdh, exhibit quite similar organ fusion phenotypes at the rosette stage. Note the differences between the serration of the wild-type leaves and those of the mutants. Leaf deformations in these mutants are not always an after-effect of a fusion at the early developmental stage. (B) Toluidine blue test in lcr, bdg and fdh. Wild-type and mutant leaves are shown after 2 min immersion in a 0.05% toluidine blue solution and destaining in water. No staining is observed with the wild type control. Bar is 1 cm. (C) Chlorophyll leaching test in lcr, bdg and fdh. Typical rosette leaves (∼ten) from seven-eight week old plants (short day conditions) were combined to both compose a sample and spectrophotometrically measure the rate of chlorophyll extraction into an ethanolic solution. The results are mean % values±standard error of at least six replicates. (D–G) Expression of the LCR:GFP reporter in transgenic Arabidopsis plants. Shown are the floral organs in which LCR:GFP expression is most conspicuous. The green color in the dual-color fluorescent confocal images corresponds to the GFP signal, and the red color to the autofluorescence of the chlorophyll. (D) The pistil, composed of the ovary (or), style (sy), and stigma covered with elongated papilla cells (pa). (E) Optical section through a pistil exposing the interior of the ovary. Note the epidermis-specific expression of GFP in the ovary wall. (F) Cross-section through the floral bud revealing epidermal GFP expression in sepals, petals, stamens and the pistil, and the GFP signal in developing pollen. The pistil epidermis exhibits a brighter signal. (G) Cross-section of the pistil showing septum (se), ovule (ov) and ovary wall in greater detail. Note that the inner epidermis of the ovary wall and the septum are devoid of the GFP signal. Bars are 200 µm for (D,E,F) and 100 µm for (G).

Figure 2. Ultrastructural aspects of the cuticle membrane in lcr and fdh rosette leaves.

The tissues were embedded, and ultrathin sections were stained, and examined through transmission electron microscopy (TEM). Plant types are indicated. (A–E) Young leaves. Adaxial (A,B,D) and abaxial (C,E) cuticle. (F–O) Adult leaves. Adaxial (F,G,H,I,J) and abaxial (K,L,M,N,O) cuticle. Note that the regular electron-dense cuticle proper in both the wild type (wt) and fdh are structurally distinct from that in lcr. In lcr, the cuticle is characterized by depositions of an electron dense crystalloid material, multiple cutin layers, cavities inside the cell wall and breaches at the surface; (O) depicts a notable solid shape bulging out of the cell wall in lcr. Bars are 200 nm.

Figure 3. Resistance of cuticular mutants to Botrytis cinerea.

(A) The appearance of leaf symptoms. Detached leaves were placed on the agar media in Petri dishes, droplet-inoculated with conidiospores of B. cinerea strain 2100 and examined at 3 dpi. Three representative leaves are shown per genotype. (B) Percentage of outgrowing lesions at 3 dpi (mean±se). At least 50 leaves were used per genotype, and the statistical significance was calculated in comparison to inoculated wild type using Fisher's exact test. Different letters indicate significance at P<0.01. All differences between mutants and wild type were significant at the P<0.001; lcr and fdh are more susceptible than lacs2 and bdg at P<0.05. (C) Surface lesion areas (mean±se) for the same samples shown in (A) and (B). Genotypes assigned the same letters are statistically similar to each other at P<0.05.

Figure 4. Analysis of leaf residual-bound lipids and wax in fdh and lcr.

(A) The fatty acid composition analysis of the leaf residual bound lipids that are left after exhaustive extraction with methanol/chloroform. Values are mean±standard error for five (lcr), six (wild type) or seven (fdh) replicates, each containing leaves from at least 15 plants. (B) The composition of the analysis of leaf wax. Values are mean±standard error from six (fdh) or seven (wild type and lcr) replicates, each containing leaves from at least ten plants. Stars indicate in (A) and (B) a significant Mann-Whitney test (two-tailed, P<0.05) for mutant versus wild type.

Figure 5. An overview of the microarray results in the three mutants.

(A) Gene categories revealed by Rank Product statistical analysis, followed by a Venn diagram representing overlapping or non-overlapping gene sets. Differentially expressed genes were defined by pfp<0.05 (corresponds to FDR<0.05) between mutant and wild type samples. The set, which comprises 87 upregulated and 2 downregulated genes common to all three mutants, is detailed in Table S3. (B) Confirmation of the microarray data by semi-quantitative RT–PCR. Thirteen candidates were compared to ACTIN2 as a control. To allow the semi-quantitative estimation of differences, the number of PCR cycles has been optimized for each gene.

Figure 6. Expression pattern of DAISY in the lcr, bdg, and wild-type plants.

The tissue-specific localization of DAISY mRNA was revealed by in situ hybridization using the antisense riboprobe. (A) Cross-section showing expression in the wild type rosette. Arrow indicates an in situ hybridization signal in a vascular bundle of developing leaves. The upper right insert shows a magnified view of a vascular bundle. (B) Cross-section through the stylar tissue of a wild-type pistil, revealing the expression signal in the pollen transmitting tract and ovules (arrows). (C) Cross-section of the lcr rosette. (D) Cross-section of the bdg rosette. Note that DAISY is ectopically induced in young developing leaves. (E) Semi-longitudinal section showing fusions between younger and older leaves in lcr. Arrow depicts signals in the vasculature of the older leaf. (F) Cross-section through fusions between younger and older leaves in bdg. Arrows depict signals in the vasculature of the older leaf. Note that the hybridization signal (E) and (F) in the younger leaves is comparable to that in the vascular bundles (arrows), and it may be somewhat stronger in the epidermis than in the inner tissues. Bars are 200 µm for (A,C,D,E,F), and 100 µm for (B) and the insert in (A).

Figure 7. Meta-analysis of differential expressed genes (DEGs) in the cell wall and cuticle-deficient mutants.

The bars illustrate gene overlaps for lcr and cesA4/irx5-differentially regulated genes when compared to a subset of DEG lists in the MASTA database (MicroArray overlap Search Tool and Analysis), which contains over 600 microarray contrasts (e.g. mutant vs. wild type, or treatment vs. control comparisons). For the best consistency, the DEGs in MASTA were selected using the Rank Product method . The top 200 upregulated and downregulated genes from each contrast were used for comparisons. The number at each bar indicates the number of genes detected in the overlap between a query and a target DEG list. Arrow points to the overlap between genes upregulated in lcr and those downregulated in se. Shown are only the numbers >9 that correspond to the level of P<7.1×10 ⁻⁰⁵ (vertical blue lines). The overlap values above this threshold were considered to be statistically significant.

Figure 8. Suppression of organ fusions, TB staining, and resistance to B. cinerea in the se lcr and se bdg double mutants.

(A) When grown under standard conditions (short-day photoperiod) in a greenhouse for four and a half weeks, se lcr and se bdg are phenotypically indistinguishable from se (se corresponds to the se-1 allele). Bars are 1 cm in (A,B,C). (B) TB staining differentiates se lcr from lcr, bdg and se bdg. (C) Percentage of outgrowing lesions at 36 and 72 hpi (mean±se) following infection with B. cinerea. At least 40 leaves were used per genotype, and the statistical significance was calculated using Fisher's exact test. Letters indicate significant differences in series between genotypes as determined by pairwise comparisons (P<0.05).